Based on our understanding of envelope-based mechanisms of humoral evasion, we have attempted to design envelope-based immunogens with these mechanisms disabled. Such modied immunogens with weakened defenses may elicit more broadly neutralizing antibodies. We have also devised protein scaffolding technologies, as a means of presenting structural mimics of the epitopes of broadly neutralizing antibodies to assist in their re-elicitation. Scaffolds can be non-homologous proteins, identified through structural searches of the entire Protein Data Bank (PDB). Alternatively, scaffolds can be homologous proteins, which are structurally similar, but antigenically distinct from the HIV-1 envelope glycoproteins. An alternative to scaffolding involves resurfacing, where the surface of a molecule, not involved in eliciting a desired response, is altered between prime and boost phases of immunization. We have also been investigating how insights from B cell ontogeny of broadly neutralizing antibodies identifies difficult steps in the elicitation process, which might be influenced by immunization. Finally, work with the fusion glycoprotein from respiratory syncytial virus (RSV) indicates that it is important to focus on a site rather than an epitope, and focus more heavily on a site that is sensitive to neutralization. Such a neutralization-sensitve site paradigm is allowing for positive vaccine results against RSV and may be useful against HIV-1.